And so Google's amusing update game continues, with the spam update rolling out.

Google is the actual spam at this stage.

Whatever they are doing with this update it will not reuslt in better search results.

@Martin I think only die-hard Google fans still believe that Google is rolling out an update that will improve search results.

@Micha, bingo. i bet this will hurt quality content sites.

google said goodbye to evaluate the website as a whole.
We are outranked by ecoms that professions are way away from our niche but they rank for very competitive terms while they sell some items as accessories for their items.

We have a item that is very IT related.
There is a ecom that sells boats and outranks us because they sell this item as accessories for only this boats.
google: IT not equal boats
Complete silly.
And our complete niche looks like this.

@matin I experience the same thing; the shop specializes in products from one country. No matter what keyword you search for, either a fake shop or a shop with completely different products comes up first.

In the news section of my niche, it's even funnier, as there's a pile of fake news that was clearly written by AI. Sometimes I wonder why I still do research; just making something up obviously works better.

I am seeing stronger traffic today after several days of somewhat tepid USA traffic. I do think that these updates kick in a couple of days before they are reported. It's not unusual to see higher traffic as we roll into September...so far this update doesn't seem to be having any major impact yet

I am seeing a big increase in searches coming in from ChatGPT some days...up 325%+. Still, it's a small fraction of my direct traffic and certainly my search traffic. Nice to have but not earth shaking yet.

Google is the actual spam at this stage.

Exactly. It's all the Googlespam that needs to go.

I am seeing a big increase in searches coming in from ChatGPT some days...up 325%+

I've seen a huge increase in bots/scrapers spoofing ChatGPT as the referrer. IP addresses used by OpenAI can be checked at: [platform.openai.com...]

Yeah, the BOTS and the DoS attacks are constant. I get DoS attacks on my home network.

As far as my windows web server, I am now spending about an hour a day blocking IP addresses. It has worked like a charm. They are getting easier and easier to identify and block. I don’t feel as though I’m playing wack-a-mole anymore. Right now, Vietnam seems to be the source of bad traffic.

@gatormark are u sf?

@saladtosser

I don’t know what “sf” means. Sorry. San Fran? No.

No worries I just had a feeling we were in the same niche and if you were you would know what sf stands for :) Guess not!

@gatomark

So you cannot access yr website because of DDOS attacks if so fair enough if not you seeing Chinese AI bots grabbing info!

Move on do not confuse zombie traffic with Vietnam traffic your website is fine it is just lack of humans visiting!

I've seen a huge increase in bots/scrapers spoofing ChatGPT as the referrer. IP addresses used by OpenAI can be checked at: [platform.openai.com...]

I checked and the IPs are actual users around the world who are clicking links out from GPT searches. At first I thought perhaps more students were using it since back in school, but just as much of the traffic is from outside of the US and from domestic. It's nice to have the traffic, but certainly not a bonanza. It's 4.5% of my current search traffic volume

[edited by: ichthyous at 4:42 pm (utc) on Aug 28, 2025]

@gatormark

I am now spending about an hour a day blocking IP addresses. It has worked like a charm. They are getting easier and easier to identify and block.

Yep, me too for the last few days I have taken a couple of hours in my early afternoon and have now got it down to under 30 mins.

At its worst one of my small, specialised widget sites bandwith increased tenfold. It wasn't causing a financial issue, simply visually horrible-looking stats and graphs, I don't do ugly :-)

I like it when in one most censored forums the mods do not move DDOS posts!

They have found this relevant!

I checked and the IPs are actual users around the world who are clicking links out from GPT searches.

@ichthyous
Interesting you are seeing real users. The last two ChatGPT-User/1.0 user hits I see (seconds ago) are from a Microsoft network (52.230.164.187 and 172.213.21.147). The visits you are getting, are they landing on info or product pages? We've been on OpenAI's product wait list a couple days after it was first announced months ago. I don't see anything in the OpenAI change log [help.openai.com...] that would account for a spike in traffic.

Our traffic from Google is in the gutter, but Google is crawling heavily and repeatedly hitting pages with different IP addresses.

So you cannot access yr website because of DDOS attacks if so fair enough if not you seeing Chinese AI bots grabbing info!

Move on do not confuse zombie traffic with Vietnam traffic your website is fine it is just lack of humans visiting!

@mosxu
I’m sorry, I didn’t quite follow the comment. But, I am blocking all nonhuman traffic aside from the big search engines. This includes bots, scrapers, DDoS, vulnerability probers, cross-site scripting, SQL injection, LDAP injection etc.

Based on my log files, I’m down to about 3-4 malicious attempts per hour.

nteresting you are seeing real users....The visits you are getting, are they landing on info or product pages? We've been on OpenAI's product wait list a couple days after it was first announced months ago. I don't see anything in the OpenAI change log [help.openai.com...] that would account for a spike in traffic.

@BigKat Rapid user adoption? Many more people are shifting to using it, especially on their phones. Some people live all day with it now. I should mention that I am seeing more searches from AI across the board, but ChatGPT is just the largest number.

I am seeing the traffic hit both info and product pages, but keep in mind that my site shows no public pricing and so I am not very similar to an ecommerce site...clients must inquire as to the cost. In some ways that means that my product pages are also seen as informational. They contain unique content and a unique description about that content which can easily be used for research purposes. I even have people emailing me to correct the text when I have an inaccuracy on the site.

I suspect some of these GPT searches hitting product pages might be coming from people clicking on images embedded in the searches. I would have to run searches myself for what I think might be the topics in order to test that. Unfortunately CHatGPT doesn't pass along any hint of the search performed.

ChatGPT, before I blocked it, gave me 0.00001% of my traffic but was responsible for 10% of the server load. It's basically a DoS attack. The only sensible thing to do with ChatGPT is to block it permanently and forever.

35% of requests to my server end up in 403 response as they are content theft bots of various types and they all, including ChatGPT, completely ignore robots.txt.

Anyone else with solid technical foundations (CWVs, accessibility/usability, TOCs, breadcrumbs and matching rich schema) seeing drops?

@Saladtosser,

we are seeing the same in our niche.
And the overall approval of the site seems offline. Means a site that is having its profession in fruits and that sells also some technical items like USB Sticks could easily outrank a site that only sells USB Sticks and give more technical inside information.
Our niche is full of examples. Never seen shops that i would never trust just because they look like backyard sales.

I don't buy this theory. Penalizing sites for good speed optimization and some honestly implemented technical SEO would be too stupid even for Google.

@christianz,

i know this sounds kind of stupid. But realize that AI Mode is counterproductive to ads. Users will not click on ads while having the answer in seconds.
What is the closest and simpliest solution?
Make the results as worse as poeple are just confident with.

Spam update = Spamers🦁 on top

Steep drop yesterday. I guess the Spam Update is finally starting to "work".

@Martin Ice Web agree completely mate! In ecoms I watch I see many ecoms I have never seen before taking the top 10 spots pushing the pervious ones that had held positions for 10 years to page 2. Services I watch one site was offline for 6 weeks since they didn't pay their hosting, the 404s never dropped out of the index or lost positions. Sites at the top are horrid right now!

ChatGPT, before I blocked it, gave me 0.00001% of my traffic but was responsible for 10% of the server load. It's basically a DoS attack. The only sensible thing to do with ChatGPT is to block it permanently and forever.

Is that sensible though? You are basically cutting off all future traffic that these AI platforms might send. It may be small, but not having any presence whatsoever might be worse. I can't accurately determine how much of an impact the visibility in AI results has on my business. They don't all need to clickthrough for it to have an impact.

For example, people pay for 'remarketing' ads...basically for the potential customer to be reminded of the brand that they have already visited. The more your brand / site / name pops up in results the more it reinforces your credibility.

AI is already here, and it will impact search results from now on...the lower search traffic is baked in, and probably going even lower from here. So the way I see it, traffic from AI search can somewhat offset that...even if not 100% compensatory. There's not much we can do, we are all being dragged along for the ride.

Is that sensible though

@ichthyous

I should have qualified that my site is purely information, not selling. If my site was selling something it would definitely not be sensible.

In my case ChatGPT is somewhat like a competitor that is using my stuff without permission to compete with me. It's the same for most info sites.

I blocked all AI bots in Cloudflare.

AI overviews appear to be an extinction-level event for most websites. Sigh...

Traffic drop intensifies. Yesterday new record low.

It looks like I have been classified as Spammer. This is actually a good thing. It gives me hope for future updates, because Spammers have done very well in 2025. Not just that - maybe my newfound status will be relayed to large AI companies and they will stop loading my server with relentless scraping.

It looks like I have been classified as Spammer.

Possibly the opposite of what you think. Your content may be so good Google is stealing it for AI O and must hide the evidence where they stole that content from. Wouldn't be the first time Google did this and won't be the last either.

Traffic drop intensifies. Yesterday new record low.

You sure you aren't being de-indexed after blocking Google's AI bots? They've said that if you block their AI bots completely then the site will be dropped from search index as well...